In recent years, many writers on economics have told the public that people have been hoarding large fractions of their respective incomes. Figures furnished by the United States Department of Commerce prove, however, that this notion that there has been hoarding is purely an illusion, conjured up by the great expansion in demand deposit volume brought about by borrowing from the banks by the United States Government. In reality, those who have had spending power have spent it rather promptly; so that, in any particular year, spending power and spending have really been almost synonymous. The facts of the case are set forth in table II.

TABLE II. Total spending by individuals compared for the prewar years with the total after-tax income of individuals

United States Survey of Current Business, June 1944, p. 13. Includes net increases in claims against insurance companies, savings bank deposits, building and loan assets, and holdings of United States savings bonds. The net increases in claims of policy holders against insurance companies have been calculated by subtracting "unpaid premiums" and "premium notes and loans" from the sum of "reserves" plus "unpaid dividends." Savings bank deposits data have been obtained by averaging the figures for the various pairs of fiscal years. The figures on United States savings bonds represent excesses of issues over redemptions. The data have been extracted from various issues of the Statistical Abstract of the United States.

Net taxes paid by individuals have been subtracted from the total income receipts of individuals, as reported in the Statistical Abstract of the United States for 1942, p. 358. The tax payments substracted are those estimated by Richard M. Bissell, of the U. S. Department of Commerce, and published in his article in Fortune entitled "The Anatomy of Public Spending."

Data furnished by the United States Bureau of the Census, when analyzed statistically, show that the people of the United States have tended to pay out each year since 1919, roughly $20,000,000,000 for nonvariable expenses. Approximately half of their remaining new spending power has gone to pay for the services rendered by the manufacturing industry. We find, therefore, that if, in any year, we subtract from the net new spending power $20,000,000,000, and divide the remainder by 2, we come surprisingly close to obtaining the figures furnished by the Bureau of the Census under the title "Value Added by Manufacture." The striking similarity between the two quantities is shown by comparing columns C and D in table III.

Anyone who stops to think about the matter knows that employers cannot afford to hire workers merely for the sake of giving the latter employment. The manufacturer hires help only when he has orders on hand or in sight. These orders depend upon the net new spending power of the public. When this spending power declines, orders fall off and he does not need so many workers. On the other hand, when spending power increases, orders for goods increase, and he needs more employees.

One of the remarkable facts revealed by study of the data furnished by the United States Census of Manufactures is that, year after year, whether times are good or times are bad, whether the Republicans or the Democrats are in power, whether employees are organized or unorganized, labor tends to get an almost constant percentage both of the gross value of output of all our factories, and

TABLE III.-Dependence of the total value added by manufacture upon the net total of new spending power accruing during the year in the United States

2 Reasons for the subtraction and division are discussed in the text.

3 Statistical Abstract of the United States, 1942, p. 885. The figures for 1935, 1937, and 1939 are slightly smaller than they should be to be exactly comparable with those for earlier years: for in the later years, the value of contract work has been subtracted. Were the figures made comparable, the errors in column E would be reduced for 1935 and 1939, and increased for 1937.

also of the value added by manufacture. That such is the case is made clear by the graphs in chart 1.

Since the total "value added by manufacture" is dependent upon the aggregate net new spending power in the Nation, and since wage workers receive an almost constant fraction of the value added by manufacture, it appears inevitable that the total "take-home pay" of all factory workers combined must depend upon the volume of net new spending power in the Nation. There is no way of escaping the fact that, if the total pay is known, the number of hours of employment which this pay will provide is determined by the average rate of pay per hour. If the rate of pay per hour were low enough, the factories of the Nation might employ 15,000,000 workers; if the rate of pay were high enough, they might be able to find jobs for only 5,000,000 workers. How very closely the volume of factory employment is determined by the volume of net new spending power in the Nation and the average hourly wage in factories is shown by table IV and chart 2. Reference to the last column of table IV shows that, in fact, when the number of hours of factory employment is estimated purely on the basis of net new spending power and average hourly earnings, the results come on the average within 62 percent of the total number of hours of employment actually given by our factories. The closeness of this correspondence seems to prove conclusively that the two forces just mentioned are the prime determinants of the volume of factory employment in the United States.

TABLE IV.—Comparison of expected with actual total volume of employment of factory-wage workers in the United States

Sum of realized national income (as estimated by the National Industrial Conference Board) and changes in the combined volume of demand deposits and money, as shown in Federal reports. Computed by the National Industrial Conference Board and quoted in the United States Survey of Current Business.

Calculated from U. S. Census data on employment in manufacturing and from the figures on the number of hours worked weekly compiled by the National Industrial Conference Board.

We see, therefore, that whether employment increases or decreases is purely a Net new spending power rises or falls. matter of whether the key ratio; that is, Average hourly earnings That such is the case is shown by the figures in table V. Clearly, whenever the key ratio rises, people go to work. When it falls, they lose their jobs.

Table VI points out in simplified form, the fact that, when net new spending power rises faster than average hourly earnings, employment increases; when, on the other hand, net new spending power declines more than average hourly earnings, unemployment appears. Chart 3 presents this fact graphically.

5

CHART 2

COMPARISON OF ACTUAL WITH EXPECTED VOLUME OF EMPLOYMENT IN ALL FACTORIES IN THE CONTINENTAL UNITED STATES

25

20

EXPECTED

TABLE V.-Cyclical changes in the ratio of national net new spending power, to average hourly earnings in manufacturing, compared with changes in the colume of factory employment

1 Sum of realized national income (as estimated by the National Industrial Conference Board) and changes in the combined volume of demand deposits and money as shown by Federal reports.

Computed by the National Industrial Conference Board and quoted in the United States Survey of Current Business.

Calculated from U. S. Census data on employment in manufacturing, and from the figures on the number of hours worked weekly compiled by the National Industrial Conference Board.

TABLE VI.-Percentage cyclical changes compared for net new spending power, average hourly earnings, and the volume of factory employment